Transgenic plants represent an innovative platform for the cost-effective large-scale production of various pharmaceutical proteins. The eventual open-field production of plant-made pharmaceuticals (PMPs) requires risk assessment to determine the potential for harm to the surrounding ecosystem. In the present study, the environmental persistence of a transgenic maize-expressed antigen, Escherichia coli heat-labile enterotoxin subunit B (LTB), was studied under laboratory conditions. To semiquantitatively monitor the persistence of LTB in soil, extraction with a high-salt, high-pH extraction buffer was optimized using the closely homologous Vibrio cholerae enterotoxin subunit B (CTB) as a test substance. The time to dissipation of 50% (DT50) of the extractable fraction of maize-expressed LTB was 4 to 15 d in pond water and 35 to 90 d in soils. Both extraction efficacy and persistence were strongly affected by the matrix type and incubation conditions. In contrast with maize-expressed LTB, the DT50 for bacterially produced LTB and CTB was less than 4 d both in pond water and soil. Although maize-expressed LTB was more stable than bacterially produced analogue, its dissipation was governed by an initial lag, which could be attributed to release from the plant material, followed by rapid decline.